Does torture work? Can we get reliable, actionable intelligence by torturing people?

That ugly debate has raised its head again after the death of Osama bin Laden. Some are claiming that waterboarding provided vital information that led to his whereabouts and ultimate demise. But are these claims accurate?

With the information available currently, there is no way of knowing for sure. It is entirely possible that other methods could have extracted the same information. The more important objection, however, is that the accuracy of the information obtained from waterboarding is not known. Maybe waterboarding led the U.S. to Osama. Good. But, it is only one data point. What if waterboarding also gave us 99 false leads? That would make enhanced interrogation methods only 1% accurate.

Without more data, it is simply not possible to declare waterboarding to be more successful at gathering authentic information than any other method. Thankfully, science can help.

A very straightforward experiment could provide a nearly definitive answer to the torture question. Critically, if you are willing to suspend ethics and morality, I would like to offer a modest proposal: a randomized controlled trial (RCT).

Normally, biomedical researchers conduct RCTs to determine the effectiveness of new treatments or experimental drugs. In our experiment, we will substitute potential life-saving therapies with waterboarding. Patients in need will be replaced with terrorists indeed.

Start with 100 terrorists. (If concerns about statistical significance arise, that number could easily be bumped up to 200 terrorists.) Finding voluntary participants could be challenging, but several potential enrollees are already waiting at Guantanamo Bay. Then, split them into two interrogation groups. The first group will constitute the “control” group. They will receive cookies and milk. The second group will be the “experimental group.” They also will receive cookies and milk, but a regularly scheduled daily waterboarding will be added, as well.

After five or ten years, all the information gathered from the two interrogation groups would be analyzed and assessed for accuracy. If the control group was more accurate, then waterboarding does not work. But, if the experimental group was more accurate, then the CIA may want to consider ordering more buckets.

Using this technique, there is no limit on the torture methods that could be tested. Sleep deprivation, constant tickling, and forced watching of Justin Bieber music videos could constitute follow-up experiments.

Of course, this project does pose some moral hazards. However, science is not in the business of dealing with pesky issues like ethics. Just ask the U.S. Public Health Service about all the African Americans they let die from syphilis in Tuskegee, Alabama.

Science could settle the torture question. How badly do you want to know the answer?